Quantizing Convolutional Neural Networks for Low-Power High-Throughput Inference Engines
This addresses the need for low-power, high-throughput inference engines in applications ranging from data centers to embedded devices, representing an incremental improvement in quantization methods.
The paper tackles the problem of high computational resource demands for deep learning inference by proposing a quantization scheme that uses more efficient arithmetic than half-precision floating-point, achieving end-to-end post-quantization accuracies comparable to the reference model with calibration from a single inference batch.
Deep learning as a means to inferencing has proliferated thanks to its versatility and ability to approach or exceed human-level accuracy. These computational models have seemingly insatiable appetites for computational resources not only while training, but also when deployed at scales ranging from data centers all the way down to embedded devices. As such, increasing consideration is being made to maximize the computational efficiency given limited hardware and energy resources and, as a result, inferencing with reduced precision has emerged as a viable alternative to the IEEE 754 Standard for Floating-Point Arithmetic. We propose a quantization scheme that allows inferencing to be carried out using arithmetic that is fundamentally more efficient when compared to even half-precision floating-point. Our quantization procedure is significant in that we determine our quantization scheme parameters by calibrating against its reference floating-point model using a single inference batch rather than (re)training and achieve end-to-end post quantization accuracies comparable to the reference model.